Re: [ferret_users] Pick up the only values at the peaks

[Jian Ma <jianma@xxxxxxxxxx>]

Hi Peter,

Have you thought about using "if" to make selections based on the temporal 
derivative? If your resolution is high enough, you can do:

let ut=u[t=@ddc]
let u1=if ut gt -1*utm and ut lt utm then u else land

where u is your variable, utm is a limit of derivative you choose, land is 
the missing value, ut is the temporal derivative and u1 is new variable. If 
it works, this may give you both the high and low peaks, you can 
additionally use the second derivative to devide them. Oh, maybe you can 
just use only the second derivative. Simply make a judgement if the second 
derivative is lower than a negative value.

If your resolution is not very high, you can consider using some modified 
derivative.

Anyway, I've never tried this method which just jumped into my mind. Pls let 
me know if it works if you wanna give a try.

Good luck,
Jian

----- Original Message ----- 
From: "Seon Tae Kim" <seonk@xxxxxxxxxx>
To: <ferret_users@xxxxxxxx>
Sent: Tuesday, October 02, 2007 2:33 PM
Subject: [ferret_users] Pick up the only values at the peaks


> Hi All
> Now since I can not access the mail archives sever for ferret help,
> I need your help.
> I have time series data of SST anomaly.
> I want to pick up the values at the peak of the time series curve.
> Good ways never come into my head.
>
> Could anyone help me with it?
> Thank you...
> -Seon Tae-